Construction and Building Materials 15 Ž2001.

93᎐103

Assessment of the durability of concrete from its permeation

properties: a review

Lulu Basheer a,U , Joerg Kroppb, David J. Cleland c

a
School of Ci¨ il Engineering, The Queen’s Uni¨ ersity of Belfast, Belfast, Northern Ireland, UK
b
Department of Ci¨ il Engineering, Hochschule Bremen-Uni¨ ersity of Applied Sciences, Bremen, Germany
c
School of Ci¨ il Engineering, The Queen’s Uni¨ ersity of Belfast, Belfast, Northern Ireland, UK

Abstract

Concrete is a versatile and most popular construction material. Its long-term performance depends on the interactions with the
service environment, in which the penetration of deleterious substances is highly significant. The latter can be considered solely
controlled by permeation properties of the near surface concrete. This paper reviews first the various transport mechanisms
which govern the ingress of deleterious substances into concrete and lists tests which could be used to determine these
mechanisms. This is followed by a discussion on various mechanisms of deterioration of concrete in an attempt to highlight the
transport mechanisms, which are relevant to each mechanism of deterioration. Finally, from the documented information the
usefulness of permeation tests to assess the durability of concrete is discussed by presenting inter-relationships between the two
obtained in laboratory studies. 䊚 2001 Elsevier Science Ltd. All rights reserved.

Keywords: Air permeability; Carbonation; Chloride ingress; Concrete; Diffusion; Durability; Freeze᎐thaw deterioration; Water absorption;
Water permeability

1. Introduction of practice dealing with the design and construction of

The unique combination of steel and concrete has
made concrete one of the most popular construction
principles in the world. However, the lack of knowledge
about the long-term performance of concrete and the
severity of environmental impacts has caused serious
problems. There are different causes of deterioration
of reinforced concrete structures such as corrosion of
reinforcement bars due to carbonation or chloride
ingress, freezing and thawing action, sulfate attack,
alkali aggregate reaction, etc. The use of wrong materi-
als or false design, improper detailing, insufficient qual-
ity control, inadequate curing, etc., have reduced the
service life of the structures or have forced extensive
repairs, with a huge economic costs. As a result, more
restrictive regulations were introduced in various codes
U
Corresponding author.
these structures. It is now known that in the new
European Standard on Concrete, the performance,
production and conformity will rely on prescriptive
specifications of minimum grades of concrete, mini-
mum binder content and maximum water᎐binder ratio
to deal with durability of concrete structures in differ-
ent, but well defined environmental classes.
The economic impact of the durability problem has
led to extensive research for over two decades and has
initiated the way to the production of better and durable
concrete or reinforced concrete structures. Many test
methods and techniques developed over the years can
be used to carry out a residual life time assessment of
existing structures or an evaluation of the expected
corrosion resistance of concrete mixes, elements or
newly built structures. Many of these methods consider
the penetration of deleterious substances into concrete
as a rate-controlling step in the degradation process.
This paper describes some of the main causes of deteri-

0950-0618r01r$ - see front matter 䊚 2001 Elsevier Science Ltd. All rights reserved.
PII: S 0 9 5 0 - 0 6 1 8 Ž 0 0 . 0 0 0 5 8 - 1
94 L. Basheer et al. r Construction and Building Materials 15 (2001) 93᎐103

oration of concrete, and points out the transport mech- equation generally referred to as Fick’s second law of
anisms involved. Then, experimental results from the diffusion, describes the change in a unit volume with
literature on the inter-relationships between them are time w5x.
presented.
The engineering properties of concrete such as ␦ cr␦ t s ␦r␦ tx Ž D ␦ cr␦ x . Ž2.
strength, durability, shrinkage and permeability are di-
Here, D may be constant or a function of different
rectly influenced and controlled by the number, type,
variables, such as time, temperature, concentration,
size and distribution of pores present in the cement
location, etc. For D s constant the solution of the
paste, the aggregate components and the interface
above equation for the boundary condition of c s cŽ0,t .
between the cement paste and the aggregate. For ex-
and the initial condition of c s 0 for x ) 0 and t s 0, is
ample, strength and elasticity of the concrete are af-
given by:
fected by the total volume of pores whereas permeabil-
ity is influenced by the pore size distribution and cont- 1r2
C s c 0 1 y erf Ž xr2 Ž Dt .
ž ./ Ž3.
inuity w1,2x. It is believed that capillary voids larger than
50 nm, referred to as macropores are detrimental to
where erf is the standard error function.
strength and impermeability, whereas voids smaller than
A factor to consider while dealing with the diffusion
50 nm referred to as micropores are more related to
process is the chemical reactions taking place between
drying shrinkage and creep w3x. The water existing in
the penetrating substances and concrete. For example,
pores bigger than 50 nm behaves as free water and it
the diffusion of chloride ions into the concrete is
plays an important role in the durability of concrete.
accompanied by reaction such as physical and chemical
binding at the hydration products. Then, Eq. Ž2. must
1.1. Transport mechanisms in concrete
be extended by a sink which itself may represent a
function of different variables. The reaction reduces
The movement of gases, liquids and ions through
the concentration of movable chloride ion at any par-
concrete is important because of their interactions with
ticular site and, hence, the tendency for inward diffu-
concrete constituents or the pore water, and thus it can
sion is further reduced. The effect of diffusion with and
alter the integrity of concrete directly and indirectly
without chemical reaction is shown in Fig. 1 w6x. In
leading to the deterioration of structures as mentioned
experiments, which do not explicitly consider binding,
above. These movements, generally called penetration,
erroneous estimates are made using the diffusion Eq.
occur due to various combinations of air or water Ž2.. Therefore, frequently an ‘apparent’ diffusion coef-
pressure differentials, humidity differentials and con-
ficient is deduced from experiments, which then de-
centration or temperature differences of solutions w4x.
pends on time t w7x.
Depending on the driving force of the process and the
1.1.1.1. Diffusion tests. The transport of either gas,
nature of the transported matter, different transport
water vapour or ion due to a concentration gradient
processes for deleterious substances through concrete
across concrete can be used to determine its diffusion
are distinguished as diffusion, absorption and
characteristic. In general diffusion tests can be classi-
permeation.
fied as:
1.1.1. Diffusion
䢇 gas diffusion test;
Diffusion is the process by which matter is
䢇 water vapour diffusion tests; and
transported from one part of a system to another due
䢇 ionic diffusion tests.
to concentration gradient. The macroscopic movement
occurs as a result of small random molecular motions,
which take place over small distances. The progress of
diffusion is much faster in gases than in liquids, solids
being the slowest. Flux is a measure of diffusion which
is the flow rate per unit area at which mass moves.
Fick’s first law of diffusion states that the rate of
transfer of mass through unit area of a section, J, is
proportional to the concentration gradient ␦ cr␦ x and
the diffusion coefficient w5x:

J s yD␦ cr␦ x Ž1.

For non-steady state conditions, the concentration c

at the location x changes with time, and the balance Fig. 1. Diffusion with and without chemical reaction w6x.
 L. Basheer et al. r Construction and Building Materials 15 (2001) 93᎐103
In the gas diffusion test two streams of gas of equal
pressure and temperature are passed along either side
of a specimen. Therefore, the transfer of gases by
diffusion is stimulated by a difference in concentration.
Traces of one gas are detected in the stream of the
second gas Žwhich is normally an inert gas. to measure
the rate of gas diffusion through the specimen w8x.
There are two different types of water vapour diffu-
sion tests. Water vapour transmission tests and water
vapour transpiration tests. In the water vapour trans-
mission test, the water vapour passing through the
specimen is collected either by condensing it or by
absorbing it with a desiccant. In water vapour transpir-
ation tests, the loss of weight of a saturated specimen
due to the evaporation of water is measured w8x, and
transport parameters may be deduced from the drying
curve.
As mentioned earlier, ionic transport in concrete can
involve chemical or physical binding. In order to de-
termine the diffusion coefficient, binding is often ne-
glected and pure diffusion, some times with binding
capacity, is adopted. Three types of ionic diffusion tests
are normally used: steady state diffusion tests, non-
steady state diffusion tests and electric field migration
tests.
In the case of a steady state diffusion test, a thin
slice of the test specimen forms a barrier between an
ion source solution and another suitable solution free
of the ion. This allows ions to diffuse due to a concen-
tration gradient. The rate of diffusion is obtained by
periodically determining the ion content in the sink
solution. When steady state conditions are achieved,
the diffusion coefficient is calculated by using Fick’s
first law of diffusion.
In the non-steady state diffusion tests, the penetra-
tion of ions is achieved either by immersing the speci-
men in a solution containing specific ions for a certain
time or by ponding the solution containing specific ions
on the test specimen for a specific duration. The pene-
tration is maintained uni-directional by sealing all ex-
cept one surface of the specimen. Either the penetra-
tion depth or the penetration profile of the specific ions
in the specimen are used for applying Fick’s second law
of diffusion and calculating the diffusion coefficient by
regression analysis of Eq. Ž3. w9,10x.
Due to very long test periods required for both
steady state and non-steady state diffusion tests, meth-
ods where the transport of ions can be accelerated by
the application of electrical potential gradients, called
electrical migration tests, have become very popular
recently. Early developments of this category of tests
did not consider the complex phenomena of ion trans-
fer satisfactorily w11x. However, recent developments
consider the similarity between diffusion and migration
by introducing the concept of the ion mobility w12x. This
has enabled the determination of an ion diffusion
95
coefficient from the migration tests. Depending on the
purpose of testing and evaluation procedure, migration
tests may simply focus on the electrical charge passing
within a present testing period w13,14x. Similar to diffu-
sion tests, the penetration depth of chloride ions in a
non-steady state test or the amount of ions passing
through the specimens in steady state conditions can
also be observed w15x.
1.1.2. Absorption
Transport of liquids in porous solids due to surface
tension acting in capillaries is called water absorption.
Absorption is related not only to the pore structure,
but also to the moisture condition of the concrete
w16᎐18x.
The absorption of water into the dry concrete is
considered to have two basic parameters.
1. the mass of water which is required to saturate the
concrete Žthe effective porosity.; and
2. the rate of penetration of the capillary rise Žthe
sorptivity..
Since the filling of capillary channels and voids and
advancing of water occur almost side-by-side during
absorption, a combined effect only can be measured,
which will give a capillary effect. Even if the test
methods are varied and the results obtained from them
appear to be different, similar equations are obtained
for water absorption behaviour w19x.
A s C q St 1r2 Ž4.
where A is the term relating to the water intake, S the
sorptivity and t the elapsed time. C is the initial
disturbance observed by some researchers and is
believed to be dependent on the surface finish.
1.1.2.1. Absorption tests. From the literature it can be
seen that many different types of equipment are used
to measure the absorption of concrete w8x. These ab-
sorption tests can broadly be divided as:
1. surface absorptivity test; and
2. drilled hole absorptivity test.
As the name indicates, for the first method, a test is
carried out on the surface of the concrete, whereas for
the second method a hole is drilled and water is admit-
ted through it. For assessing the absorption behaviour
of concrete, different methods are adopted for differ-
ent equipment. In some cases, the flow of water over a
total elapsed time is measured, whereas in some other
cases, the inflow of water is continuously monitored
over a specific duration of the test w8x. The Standpipe
Absorptivity test, the Initial Surface Absorption test,
and the Autoclam Sorptivity test are three typical sur-
96 L. Basheer et al. r Construction and Building Materials 15 (2001) 93᎐103
face absorptivity tests w8x. Tests in the category of
drilled hole absorptivity tests include the Figg water
permeability test w20x and the Covercrete Absorption
test w21x.
1.1.3. Permeability
Permeability is defined as that property of a fluid
medium which characterises the ease with which a fluid
will pass through it under the action of a pressure
differential. Darcy’s law states that the steady-state
rate of flow is directly proportional to the hydraulic
gradient, i.e.
¨ s QrAs yK Ž dhrdL.
where ¨ is the apparent velocity of flow, Q is the flow
rate, A is the cross-sectional area of flow, I is the
hydraulic gradient and dh is the head loss over a flow
path of length dL. K is called the coefficient of perme-
ability. Darcy’s law has been generalised to apply to
any fluid flowing in any direction through a porous
material, so long as the conditions of flow are viscous.
The law can be expressed by the equation:
¨ s QrAs y Ž kr␮ .Ž dPrdL.
where dP is the pressure loss over the flow path dL, ␮
is the viscosity of the fluid and the constant k is
referred to as the intrinsic permeability of the porous
medium. Intrinsic permeability, with the dimensions of
area, is the most rational concept of permeability, as it
depends purely on the characteristics of the porous
medium and is independent of those fluid characteris-
tics which govern the flow, i.e. viscosity ␮ expressing
the shear resistance of the fluid. For gases, the com-
pressibility must be considered as well. Then, the pres-
sure difference is related to the mean pressure w22x.
1.1.3.1. Permeability tests. For measuring permeability,
the most commonly used method is a permeability cell.
Permeability cells of various specifications and dimen-
sions have been employed in order to admit fluid under
pressure to one side of the specimen and measure the
flow either at the inlet or at the outlet w8,23x. For
steady state conditions a value of permeability is calcu-
lated from the knowledge of sample geometry and fluid
characteristics and the measurement of flow rate and
the applied pressure. However, if a steady state condi-
tion cannot be established due to either low permeabil-
ity of the test specimen or limitations of the test
conditions, a non-steady state of flow occurs. Some
experimental set-ups also do not provide exact informa-
tion on the flow paths, thus dPrdL is not defined. In
these cases a permeability index is derived. The Figg
air permeability test, Hong and Parrott’s test, Hilsdorf’s
test, and the Autoclam air permeability test are some
of the permeability tests available w23x. Among these
tests the Autoclam air permeability test can be used for
high performance concrete and surface treatments for
which the flow rates are very low.
2. Causes of deterioration of concrete
The ingress of various ions, liquids and gases from
the environment is responsible for the deterioration of
concrete directly or indirectly. For instance, the ingress
of chlorides or carbon dioxide would depassivate the
steel in concrete, and in the presence of oxygen and
water, steel may start corroding. Similarly, the ingress
Ž5. of chemicals, such as acids, alkalies and sulfates are
responsible for the chemical deterioration of concrete.
Moisture movement during freezing and thawing action
also causes deterioration of concrete. Some important
causes of deterioration are explained in the following
section.
2.1. Carbonation
When carbon dioxide diffuses into concrete, in the
presence of water, it reacts with calcium hydroxide to
Ž6. form calcium carbonate. As a result, the pH of the pore
solution is reduced to below 10. If the carbonation
front progresses until the surface of the reinforcement
is reached, the passive layer of the steel surface will be
dissolved and the corrosion of the reinforcement will
start. Carbon dioxide also reacts with aluminates and
calcium silicate hydrates, i.e. the major constituents of
hydrated cement, CSH gel, is subject to carbonation.
The gel is decomposed into calcium carbonate and an
amorphous silica gel with a porous structure w24x.
The products of carbonation alter the pore structure,
the permeability of concrete and the chemical reactiv-
ity of the hydrated cement paste matrix. The binding
capacity for aggressive ions are also affected due to
carbonation. Chemically combined and physically ad-
sorbed chloride ions are set free with the advance of
carbonation, increasing the number of free chloride
ions in concrete. This interaction between carbonation
and chloride contamination of concrete is the likely
cause of the most severe corrosion problems encoun-
tered in practice w25x.
Carbonation rate is controlled by the ingress of car-
bon dioxide into the pore system by diffusion, with a
concentration gradient of carbon dioxide acting as a
driving force w26x. The diffusion of carbon dioxide al-
ways takes place through concrete sections, which are
already carbonated. It is usually assumed that the reac-
tion front progresses after all alkaline material has
been transformed. Therefore, the carbonation rate is
diffusion-controlled and the diffusion coefficient for
carbon dioxide in carbonated concrete is the character-
istic transport coefficient. Assuming a constant diffu-
 L. Basheer et al. r Construction and Building Materials 15 (2001) 93᎐103
sion coefficient for the carbonated layer, the depth of
carbonation can be derived from Fick’s first law of
diffusion:
X s C't
where x s depth of carbonation at time t Žm., t s time
Žs., and C s constant.
However, the application of Fick’s law of diffusion
using constant parameters has limitations because D
varies with location x and strongly depends on the
moisture content in the concrete pores. For older con-
crete, the depth of carbonation observed is often less
than expected according to the 6t relation. Further-
more, concrete surfaces exposed to rain show lower
carbonation rates than those sheltered from rain.
Therefore, time functions xs An with n - 0.5 have
been proposed w27x.
2.2. Chloride ingress and corrosion due to chloride ingress
There are interactions between chloride ions and
hydration products of the cement, which may con-
tribute to the development of frost damage w28x. How-
ever, the fundamental destructive effect of chlorides is
their influence upon the reinforcement corrosion
process, primarily due to their capacity to negate the
corrosion inhibiting properties of the alkaline cement
paste pore solution w29x. This risk increases with in-
creasing concentration of free chlorides in the pore
solution w30,31x. It is generally believed that there is a
threshold concentration of the chloride ions, which
must be exceeded before corrosion occurs. The thresh-
old concentration may depend on the concrete compo-
sition and on environmental parameters w25x
The corrosion due to chloride ingress progresses at a
much higher rate than that due to carbonation. In
extreme cases, the corrosion rate in real structures can
be 5 mmryear compared to 0.05 mmryear for carbona-
tion-induced corrosion w32x. The corrosion can be ei-
ther pitting corrosion, in which a very small anode is
surrounded by a large area of cathode, or general
corrosion with a homogeneous distribution of micro
anodes and cathodes. General corrosion may have more
total loss in iron, but pitting corrosion causes more loss
in cross-sectional area and hence is more dangerous.
The action of concrete containing chloride is very
complex. The tricalcium aluminate and to a lesser
extent, calcium alumino ferrites in the cement combine
with chloride ions forming calcium chloroaluminate,
most commonly known as Friedel’s salt and chlorofer-
rite hydrates. These reaction products are in equilib-
rium with free chlorides in the pore water. The fixed
chlorides do not contribute to the corrosion process.
However, the fixing of chlorides is not permanent but
97
depends on factors such as leaching and alkalinity of
the concrete w33x.
Different transport mechanisms such as the
permeation of a salt solution, the capillary absorption
Ž7. of chloride-containing liquids and the diffusion of free
chloride ions can be associated with chloride ingress in
concrete. When capillary pores are relatively dry, ab-
sorption dominates and when they are relatively satu-
rated, diffusion becomes the dominant transport
process. These mechanisms may act simultaneously or
they may prevail in sequence during consecutive peri-
ods of time. Similarly, as per the location within the
concrete the transport mechanism can be different. If
chloride solution penetrates through the near-surface
layer by non-steady permeation, at the penetration
front capillary forces become predominant along the
coarse capillaries and the diffusion of free ions may
then lead to a homogeneous distribution of the ions
into remote sites of the paste.
2.3. Freeze᎐thaw deterioration
The deterioration of concrete under the action of
freezing and thawing is a very complex phenomenon
and many theories have been presented to explain
individual aspects of damaging effects. Among the most
important actions, which may lead to disintegration of
concrete, the following mechanisms must be considered
w28,30,34x:
1. Generation of hydraulic pressure due to freezing in
capillaries.
2. Diffusion of gel water into capillaries followed by
freezing.
3. Differential strains due to localised shrinkage and
swelling as well as thermal strains.
4. Osmotic pressures resulting from partial freezing in
capillaries of solutions with a local salt concentra-
tion.
For the development of frost damage it is essential,
however, that a high degree of saturation of the capil-
laries with water is prevailing.
The ease with which water can move into and within
concrete depends on the pore structure characteristics
of the concrete. According to Powers w35x the destruc-
tive stress is produced by the restricted flow of dis-
placed water away from the region of freezing, the
pressure being due to the viscous resistance to such
flow through the permeable structure of concrete. Al-
though the pore size distribution in concrete mainly
dictates the extent of damage to concrete during
freeze᎐thaw cycles, the amount of freezable water pre-
sent in the capillary pores is also important. The amount
of freezable water depends on the degree of saturation,
98 L. Basheer et al. r Construction and Building Materials 15 (2001) 93᎐103

the minimum temperature reached by the water and

the composition of the pore solution w36x.
Since the resistance to flow at a given rate is propor-
tional to the length of flow path, it follows that there is
a critical length of flow path, beyond which the hy-
draulic pressure exceeds the tensile strength of the
material. The internal tensile stress developed due to
freezing is reported to be in the range of 1᎐4 Nrmm2
w31x. Therefore, internal microcracks may develop ini-
tially and repeated cycles of freezing and thawing may
result in permanent damage to the concrete. According
to Fagerlund w36x, there is a critical moisture content
for each concrete at which the concrete will start to
deteriorate under freezing and thawing. Whether this
moisture content will be achieved depends on the expo-
sure conditions of a concrete element and its transport
characteristics for water, i.e. its absorption behaviour.
3. Inter-relationship between permeation properties
and durability of concrete
Different standard test methods can be used to mea-
sure the transport properties relevant to the various
mechanisms of deterioration. Experimental evidence
exists illustrating the correlation between the relevant
transport properties and either the penetration of dif-
ferent aggressive substances or the mechanisms of de-
terioration. The following section discusses the inter-
relationship between the transport mechanisms and the
above presented deterioration processes.
3.1. Carbonation
The carbonation of concrete has been correlated
with the gas permeability and water absorption by
different researchers. This is because that the tests
characterise the pore structure of the surface near
sections of the concrete and give an indirect measure
of the open and continuous capillary porosity.
Fig. 2. Carbonation depths at 1.5 years vs. water absorption after 1.5
years of exposure w37x.
and its depth of carbonation after 1 year storage in a
controlled atmosphere of 20⬚C and 65% RH with natu-
ral CO 2 content was studied by Hilsdof et al. w39x. Fig.
4 shows the correlation between the depth of carbona-
tion of concrete and the air permeability for different
types of cement. This indicates that a linear relation-
ship may be established if the square of the depth of
carbonation is plotted against the logarithm of the air
permeability. Fig. 5 gives the relationship between car-
bonation depth and the Autoclam carbon dioxide
permeability index w19x, which again gives an almost
linear correlation between the depth of carbonation
and the log of the gas permeability values. Fig. 6 shows
a similar result for the Figg air permeability test as
well. It can be seen in all these cases that there is a
good correlation between the depth of carbonation and
the gas permeability value. However, the precise re-
gression relationship depended on the method of test
used to determine the gas permeability value, i.e. the
individual transport characteristics ‘air permeability’,
‘Autoclam permeability index’ and ‘Figg air permeabil-
ity’ are not directly comparable on a quantitative basis.

3.1.1. Water absorption

Investigations carried out by Parrott w37x showed that
the depth of carbonation after 1.5 years of exposure,
both in the lab and on site can be related, to the water
absorbed after 4 h of wetting ŽFig. 2.. Dhir et al. w38x
obtained a reasonably good correlation between the
depth of carbonation and the 10-min initial surface
absorption value ŽFig. 3.. This implies that the short-
term absorption gives an indirect information on the
continuity of the pore system in the near surface zone,
which in turn is vulnerable to carbonation if very
permeable.

3.1.2. Air permeability Fig. 3. Relationship between carbonation depth and water absorp-
The correlation between air permeability of concrete tion w38x.
 L. Basheer et al. r Construction and Building Materials 15 (2001) 93᎐103
Fig. 4. Relationship between depth of carbonation after 1 year and
air permeability at an age of 56 days w39x.
3.2. Chloride ingress and corrosion due to chloride ingress
As mentioned earlier, the corrosion of reinforcement
embedded in concrete of high alkalinity will take place
only after the chloride concentration at the level of
bars exceeds a certain threshold concentration. For the
assessment of corrosion protection of the reinforce-
ment, either concentration profiles for chlorides or the
rate of progress of a given chloride concentration into
the concrete section are of interest.
3.2.1. Diffusion
Several investigators have determined an apparent
diffusion coefficient from an achieved chloride profile
in concrete after exposure to chloride solutions for a
certain period of time. In the work done by Basheer et
al. w41x this method and a chloride diffusion cell with an
applied voltage have been used. A range of water᎐ce-
ment ratios varying from 0.4 to 0.7 and aggregate᎐ce-
ment mass ratios from 3 to 8 were used for the study.
Fig. 5. Relationship between carbonation depth and Autoclam car-
bonation permeability index w19x.
99
Fig. 6. Relationship between depth of carbonation after 20 weeks of
accelerated carbonation and Figg air permeability index w40x.
Fig. 7 gives a correlation between the carbonation
depth and the chloride diffusion co-efficient from the
same study, which shows that there exists a good rela-
tionship between different transport mechanisms such
as gas diffusion and ionic diffusion.
3.2.2. Water absorption
Basheer et al. w42x report results of a study carried
out relating absorption characteristics of concrete to
chloride-induced corrosion of steel in concrete. A
weekly cyclic ponding regime was used to allow the
transport of chlorides into the concrete and the tests
lasted for 44 weeks. The results reported here are from
three water᎐cement ratio concretes Ž0.45, 0.55, 0.65.
with the same mix proportion of other constituents.
Fig. 8 shows the relationship between sorptivity index
obtained by Autoclam permeation system and water
soluble chloride ion content at 25 mm and 40 mm
depth from the surface of ponding. It shows that a good
linear relationship exists in the case of 25-mm depth of
cover. However, as the cover depth also seems to
influence the chloride content, predicting chloride con-
centration at a certain depth from the water absorption
tests can be difficult. To compare the expected protec-
Fig. 7. Relationship between chloride diffusion coefficient and car-
bonation depth for basalt concrete w41x.
100 L. Basheer et al. r Construction and Building Materials 15 (2001) 93᎐103
Fig. 8. Relationship between sorptivity index and chloride ingress
w43x.
tive abilities of different concretes water absorption
tests can be used as a quick indicative test.
Fig. 9 shows the relationship between the time to
initiation of corrosion and the sorptivity of the near
surface concrete. In the case of 25-mm cover depth
concrete, a good linear relationship exists between
Autoclam sorptivity index and the time to initiation of
corrosion w43x as in the case of chloride content in Fig.
8. In the case of 40-mm cover depth, reinforcement
bars were not corroding at the end of the 44 weeks test
for the 0.45 water᎐cement concrete. This part of the
graph is shown as a dotted line in the figure and it is
possible that the relationship between sorptivity and
corrosion initiation time can be more complex for a
deeper cover depth.
3.3. Freeze᎐thaw deterioration
3.3.1. Water absorption
Fagerlund w36,44x has developed a testing concept in
which the critical degree of saturation serves as a
Fig. 9. Relationship between sorptivity index and corrosion initiation time w43x.
criterion to evaluate the potential frost resistance of
concrete. The critical degree of saturation is the mois-
ture content at which marked damage occurs to con-
crete during the freeze᎐thaw test. According to this
concept, if the critical degree of saturation is known
from the capillary absorption rate it should be possible
to predict the number of cycles after which this satura-
tion level will be reached.
Hilsdorf et al. w39x have reported results of their
study relating weight loss of concrete cubes after 60
cycles of freeze᎐thaw cycles. Fig. 10 shows that a good
relationship exists between water absorption and the
weight loss during the freeze᎐thaw test. Ordinary Port-
land cement and Portland blast furnace slag cement
with a slag cement of 65% by mass were used for the
study. Basheer et al. w19x obtained a very good correla-
tion between the Autoclam water permeability index
and the weight loss due to freezing and thawing ŽFig.
11.. Fig. 12 shows the correlation between the number
of freeze᎐thaw cycles at failure to ISAT 10-min value
w38x.
3.3.2. Air permeability index
Hilsdorf carried out air permeability tests before
exposing the samples to freezing and thawing tests and
the weight loss during the freeze᎐thaw test was com-
pared with the coefficient of air permeability w39x. Fig.
13 shows that there exists a linear relationship between
the two, for both ordinary Portland cement and the
Portland blast furnace slag cement. The results of the
oxygen permeability and freeze᎐thaw resistance of dif-
ferent concretes are presented in Fig. 14 w45x. The
weight losses recorded are after 100 cycles of freezing
and thawing. Fig. 15 shows the result of salt scaling test
carried out as per RILEM test w46x on gravel concrete
of different water᎐cement ratios and aggregate᎐ce-
ment ratios. The log of Autoclam air permeability
 L. Basheer et al. r Construction and Building Materials 15 (2001) 93᎐103 101

Fig. 13. Relationship between water loss after 60 cycles of freezing

and thawing vs. air permeability of concrete w39x.
Fig. 10. Relationship between freeze᎐thaw resistance and water ad-
sorption w39x.

Fig. 14. Relationship between weight loss after 100 cycles of freezing
Fig. 11. Relationship between Autoclam water permeability index and thawing vs. air permeability w45x.
and weight of scaled material for gravel concrete w19x.

Fig. 12. Relationship between freeze᎐thaw resistance and water ab- Fig. 15. Relationship between air permeability index and weight of
sorption w38x. scaled material for gravel concrete w41x.
102 L. Basheer et al. r Construction and Building Materials 15 (2001) 93᎐103
index vs. the log of cumulative scaling weight gave a
very good linear relationship.
4. Conclusions
Different transport mechanisms, such as water ab-
sorption, permeability and diffusion and the general
principles adopted for their test methods were dis-
cussed in the paper. The durability parameters such as
chloride ingress, corrosion due to chloride ingress, car-
bonation and freeze᎐thaw deterioration were also dis-
cussed. From the reported results it is evident that all
the transport processes are inter related. Also from
different investigations it has been demonstrated that
the above mentioned transport processes are important
physical properties of concrete in relation to the dura-
bility parameters discussed. Although the exact rela-
tionship varies according to the test methods adopted,
a general trend can be established from the studies
carried out. Thus the durability of concrete can be
assessed by the measurement of concrete transport
processes. A dominant role in the degradation is often
taken by the surface near sections or the concrete
cover of the reinforcement. Therefore, test methods for
the characterisation of the surface zone are of particu-
lar interest, especially in view of non-destructive evalu-
ation of existing structures.
References
w1x Powers TC, Copeland LE, Mann HM. Capillary continuity or
discontinuity in cement pastes. J Portland Cement Assoc Res
Dev Lab 1959;1Ž2.:38᎐48.
w2x Verbeck GJ. Pore structure ᎏ significance of tests and proper-
ties of concrete and concrete making materials. ASTM Special
Tech Publ 1982;169A:211᎐19.
w3x Mehta PK, Monteiro PJM. Concrete structure, properties and
materials. New Jersey: Prentice Hall, 1993:548.
w4x Nilsson LO et al. HETEK, Chloride penetration into concrete,
State-of-the-art, transport processes, corrosion initiation, test
methods and prediction models, The Report No. 53, Road
Directorate, Copenhagen, 1996:151 pp.
w5x Crank J. Mathematics of diffusion. Oxford: Clarendon Press,
1956.
w6x Verbeck GJ. Mechanisms of corrosion of steel in concrete,
corrosion of metals in concrete. ACI SP-49, 1987:211᎐19.
w7x Garboczi EJ. Permeability, diffusivity and micro-structural
parameters. A critical review. Cement Concr Res, 20 pp,
1990:591᎐601.
w8x Basheer PAM. A brief review of methods for measuring the
permeation properties of concrete in-situ. Build Struct Inst
Civil Eng 1993;99Ž1.:74᎐83.
w9x Petersen CG. RCT profile grinding kit for in-situ evaluation of
the chloride diffusion coefficient and the remaining service life
of a reinforced concrete structures, Symposium ‘Chloride
ingress in concrete structures’, Chalmers technical university,
Gothenburg, Sweden, 1993.
w10x Poulsen E. The chloride diffusion characteristics of concrete.
Approximate determination by linear regression analysis.
Nordic Concr Res ᎏ Nordic Concr Fed 1990;9:124᎐33.
w11x Andrade C, Sanjuan MA, Alonso C. Measurement of chloride
diffusion coefficient from migration tests, Paper 319, Corrosion
93, The NACE Annual Conference and Corrosion Show,
1993:11 pp.
w12x Andrade C. Calculation of chloride diffusion coefficients in
concrete from ionic migration measurements. Cem Concr Res
1993;23:724᎐42.
w13x ASTM. Standard test method for electrical indication of con-
crete’s ability to resist chloride ion penetration, ASTM C
1202-94. Book of ASTM Standards. Concrete and Aggregates,
04. 02. Philadelphia: American Society for Testing and Materi-
als, 1994:620᎐5.
w14x AASHTO Designation T277-83. Standard method of test for
rapid determination of the chloride permeability of concrete,
American Association of State Highway and Transportation
Officials, Washington, DC, 1993.
w15x Geiker M, Grube H, Luping T, Nilsson LO, Andrade C.
Laboratory test methods, performance criteria for concrete
durability. RILEM Rep 1995;12:213᎐57.
w16x Glanville WH. The permeability of Portland cement concrete.
Build Res Tech Pap 1931;3:62.
w17x Mercer LB. Permeability of concrete ᎏ 1: theoretical con-
siderations, laboratory methods, details of experimental work
with new apparatus, The Common Wealth Engineer,
1945:349᎐57.
w18x Rose DA. Water movement in unsaturated porous materials.
RILEM Bull 1965;29:119᎐24.
w19x Basheer PAM. ‘Clam’ permeability tests for assessing the dura-
bility of concrete, PhD Thesis, The Queen’s University of
Belfast, 1991:438 pp.
w20x Figg JW. Methods of measuring the air and water permeability
of concrete. Mag Concr Res Lond 1984;25Ž85.:213᎐219.
w21x Dhir RK et al. Near surface characteristics of concrete: assess-
ment and development of in-situ test methods. Mag Concr Res
Lond 1987;39Ž141.:183᎐95.
w22x Klinkenberg LJ. The permeability of porous media to liquids
and gases, drilling and production practice. New York: Ameri-
can Petroleum Institute, 1941:200᎐14.
w23x Concrete Society. Permeability testing of site concrete ᎏ a
review of methods and experience, Report No. 31. The Con-
crete Society, 1987.
w24x RILEM. International Symposium on carbonation of concrete,
Cement and Concrete Association, Wexham springs, Slough,
UK, 1976.
w25x Schießl P, editor. Corrosion of steel in concrete, report of
RILEM report Technical Committee 60-CSC, Chapman &
Hall, London, 1988.
w26x Meyer A, Wierig H, Husman K. Karbonatisierung von Schwer-
beton, Schriftenreihe des Deutschen Ausschusses fur Stahlbe-
ton, Heft, Berlin, 1967.
w27x Smolczyk HG. Written discussion Žabout carbonation evalua-
tion with time on different cements., in Proc. 5th International
Symposium Chemistry of Cements, Vol. III, Tokyo, 1968.
w28x Cordon WA. Freezing and thawing of concrete ᎏ mechanisms
and control, monograph 3. American Concrete Institute, 1966.
w29x Newman K. Common quality in concrete construction, Con-
crete International. Detroit: American Concrete Institute,
1986:37᎐49.
w30x Washburn EW. The dynamics of capillary flow. Physics Rev
XVII 1921;3:273᎐83.
w31x Bjegovic D et al. Theoretical aspects and methods of testing
concrete resistance to freezing and de-icing chemicals, SP 100,
Katharine and Bryant Mather international conference, Con-
crete Durability. ACI Detroit 1987;1:947᎐71.
w32x Arup H. Corrosion of reinforcement steel in concrete ᎏ an
overview, failure and repair of reinforced structures. London:
London Press Centre, 1984.
 L. Basheer et al. r Construction and Building Materials 15 (2001) 93᎐103 103

w33x Neville A. Corrosion of reinforcement. Concrete June w41x Basheer PAM, Henderson G, Basheer L, Long AE. Establish-
1993:48᎐9. ing the influence of interactive parameters on design for dura-
w34x Plum DR. Concrete attack in an industrial environment. Concr, bility, Report No. SMRG-21-1998, 1998:221.
J Concr Soc Lond 1984;11:8. w42x Basheer L, Cleland DJ, Long E. Protection provided by surface
w35x Powers TC. Freezing effects in concrete, Durability of con- treatment against chloride induced corrosion. Mater Struct
crete. ACI SP-47, 1975:1᎐12. 1988;31:459᎐64.
w36x Fagerlund G. Critical moisture contents at freezing of porous w43x Basheer L. Assessment of the durability characteristics of sur-
materials, Second CIBrRILEM symposium on moisture prob-
face treated concrete, PhD Thesis. The Queen’s University of
lems in buildings, Bouwcentrum, Rotterdam, The Netherlands,
Belfast, 1994:296.
1974:1᎐17.
w44x Fagerlund G. The international Cooperative test of the critical
w37x Parrott LJ. Water absorption of cover concrete. Mater Struct
1992;25Ž149.:284᎐92. degree of saturation method of assessing the freezerthaw
w38x Dhir RK, Hewlett PC, Byars EA, Bai JP. Estimating the resistance of concrete. Mater Struct 1977:10Ž58..
durability of concrete in structures, concrete. J Concr Soc Lond w45x Graf H, Bonzel J. Uber den Einflub der Porositat des er-
1994;28Ž6.:25᎐30. harteten Betons auf seine Gebrauchseigenschaften. Beton, Heft
w39x Hilsdorf HK, Schonlin K, Burieke F. Dauerhaftigkeit von Beto- 7, 1990.
nen, Institute for Massivbau and Baustofftechnologie, Univer- w46x Setzer MJ, Fagerlund G, Janssen DJ. CDF test ᎏ test method
stat Karlsruhe, 1992. for the freeze᎐thaw resistance of concrete-tests with sodium
w40x Dhir RK, Hewlett PC, Chan YN. Near surface characteristics chloride solution ŽCDF.. Mater Struct 1996;29:523᎐8.
of concrete: prediction of carbonation resistance. Mag Concr
Res 1989;41Ž148.:122᎐8.